The US Court of Appeals for the District of Columbia Circuit has been a consistent check on the Food and Drug Administration's (FDA) attempts to extend its authority. Now, it seems that the federal courts in New York are becoming a favorite venue to force the FDA to use its authority. The latest of these is an order from the US District Court for the Southern District of New York in Natural Resources Defense Council v. United States Food and Drug Administration (NRDC v. FDA), that requires the FDA to reissue a notice of withdrawal of approval for the subtherapeutic use of the antibiotics penicillin and tetracycline in food animals, and to hold hearings for the relevant drug sponsors as part of mandatory withdrawal proceedings.
The opinion itself is rather curious and arguably misunderstands the Federal Food, Drug and Cosmetic Act's (FDCA) requirements for withdrawal of drug approvals and its overall approach to risk assessment. Nonetheless, it may serve as a useful spur to better antibiotic use and scientific investigation of the very real problem of antibiotic resistance in both animals and humans.
Antibiotic resistance has been a recognized problem since the late 1960s, and the FDA has struggled with how best to regulate antibiotic use in animals almost since that time. Stated simply, antibiotic resistance occurs when bacteria mutate and become less responsive, and sometimes non-responsive, to antibiotic treatment. To some degree this is a natural evolutionary process, but there are also significant concerns that improper human and animal use of antibiotics hastens and worsens that process. Recently, mostly in hospital settings, we have faced the specter of bacteria that are resistant to virtually all existing antibiotics. If we are unable to tackle the problem, we will face prolonged human illnesses and increased deaths.
Today, the use of antibiotics in food-producing animals, such as beef cattle, pigs, poultry and sheep, is prolific. These antibiotics are administered through feed, drinking water and injections. Some of these drugs are available over-the-counter (OTC) without a prescription from a veterinarian. Much use of antibiotics in animals is for the treatment, control and prevention of disease (treatment use) but a significant part is for uses such as increasing weight gain or for improving feed efficiency (production use). The concern is that the production use of these drugs in animals will lead to greater levels of resistance that are transferred via food consumption to the human population. The full extent of the human health risk posed by that transfer is an issue of some controversy. A full understanding of the mechanism of causation is still elusive — most findings of risk are correlative rather than causal.
In 1977, the FDA announced its intent to withdraw approval for the antibiotics penicillin and tetracycline for certain livestock uses. The FDA never took subsequent action on those notices. Soon after they were issued, several congressional committees recommended that the FDA conduct research on the human health risk posed by production use before withdrawing approval of the drugs for that use. The FDA acquiesced and contracted with the National Academies for the first of many studies that have been undertaken over the last thirty-five years. That study recommended additional epidemiological research. Subsequent studies have neither been able to document that production use of antibiotics is safe nor that there is direct evidence of quantifiable risk. Unfortunately, although there are many expert reports, few of those reports contain comprehensive scientific data.
In 2003, the FDA itself concluded that the risks of antibiotic use in livestock were neither proved nor disproved, but it did not deny that there was some degree of risk, and could not conclude that continued production use of penicillin and tetracycline in food animals was safe. Guidance for Industry (GFI) #152 [PDF] sets up the basic framework that is still used by the FDA today. It applies a risk-based assessment process for evaluating new antimicrobial animal drug applications. That assessment requires consideration of the importance of the drug in human therapy, whether the use is selectively targeted or broadly applied to entire flocks or herds and the degree of veterinarian involvement in the decision to use the drugs. Drugs approved after 2003 must generally meet these standards for approval. But GFI #152 does not provide a satisfying framework for dealing with antibiotics that were approved prior to 2003. Since those antibiotics are often no longer subject to patent protection, they may be the most cost-effective options for both human and animal use.
In June 2010, after continued discussions with industry participants and others, the FDA issued a draft guidance: The Judicious Use of Medically Important Antimicrobial Drugs in Food-Producing Animals (DGFI 2010). The draft guidance uses the general framework set out in GFI #152 but adds an additional concept: "judicious use." DGFI 2010 defines "judicious" as the avoidance of any uses that are not necessary or appropriate, and concludes that antimicrobial drugs that are important for therapeutic use in humans should not be used for growth promotion and other production purposes in animals. Further, it recommends increased veterinary oversight, setting the stage for more limited OTC antibiotic use.
The court's opinion in NRDC v. FDA, filed in March 2012, centers on the legal effect of notices issued in 1977 by the FDA when it announced its intent to withdraw approval for the antibiotics penicillin and tetracycline for certain livestock uses. As noted earlier, the FDA never took any subsequent action on those notices, including refusing to take action on citizen petitions based on those notices, but it did not withdraw them either until the case was pending. Plaintiffs argued, however, and the court agreed, that the notices constitute a "finding" that penicillin and tetracycline are not safe and that finding compels the FDA to begin withdrawal proceedings.
The court acknowledged that to withdraw approval of these drugs, the FDA bears the initial burden of producing evidence "that experience or scientific data show that such drug is unsafe for use under the conditions of use upon the basis of which the application was approved." The FDA has interpreted this to mean that it must show that "there is a reasonable basis from which serious questions may be inferred about the ultimate safety of the drug and any substance that may be formed in or on food as a result of use of such drug under approved conditions." The court, however, found that the language of §512(e)(1) of the FDCA (21 USC §360b(e)(1)) is unambiguous and rejected the qualifications that the FDA sets out in its interpretation. Instead, the court held that once the FDA makes a finding that a new animal drug "has not been shown to be safe" it is required to implement withdrawal proceedings. In essence, what the court held is that once such a finding is made, it triggers proceedings that cannot be stopped regardless of the FDA's comfort in meeting its burden — even if new evidence or data is revealed that might make the FDA question its original decision.
The court maintained that such an interpretation is dictated by the statutory language of the relevant section of the FDCA. But the statutory language of §512(e)(1) does not appear as unambiguous as the court held. Reading §512(e) as a whole appears to give the secretary considerable discretion to "evaluate" the "experience or scientific data" and to consider evolving information. In addition, withdrawal proceedings are typically administratively costly and time-consuming. As a result, in many situations involving drug approvals the FDA has sought the sponsor's voluntary withdrawal or labeling changes rather than instituting formal proceedings. This appears to take that option away. Moreover, the court's interpretation could have dangerous effects — it would make the FDA extremely reluctant to make public preliminary findings about potential risks lest it find itself riding a train it cannot stop.
In addition, implicit in the court's ruling is the idea that if the FDA finds that a drug poses potential risks it has, therefore, "not been shown to be safe" and must be withdrawn. But the FDCA does not, with the exception of some substances that are shown to cause cancer, apply a standard of zero tolerance for risk. Instead, the FDA action requires evidence of causation and scientific risk assessments and — for better or worse — for a remarkably long time that evidence was murky. The FDA has withdrawn approval for only one antibiotic, enrofloxacin (marketed as Baytril), for which approval for production use was withdrawn in 2004. The FDA's latest guidance on the subject, The Judicious Use of Medically Important Antimicrobial Drugs in Food-Producing Animals [PDF] (GFI #209), issued in April this year, does provide a literature review demonstrating increasing evidence that production use of antibiotics in livestock animals is an important piece of the problem of antibiotic resistance. However, it is not clearly sufficient to justify mandatory withdrawal for all relevant antimicrobials — although it is likely enough to induce negotiations with sponsors for voluntary withdrawal.
In response to (or at least hastened by) the court's opinion in NRDC v FDA, the FDA has taken a multi-pronged approach. First, in GFI #209, the FDA explicitly continues to reject the court's interpretation of its requirement to pursue withdrawal proceedings under FDCA §512(e). But it then takes steps to increase its oversight of antibiotic use in animals and to further limit use. As noted earlier, GFI #209 provides an expansive scientific literature review which provides the foundation for the FDA's increased oversight. It also finalizes the judicious use approach proposed in DGFI 2010. GFI #209 thus adds two principles to the guidance framework set out in GFI #152. It states that use of medically important antibiotics in food producing animals should be limited to assuring animal health and that those decisions should include veterinary oversight or consultation. In addition, two days after finalizing GFI #209, the FDA issued a draft guidance: New Animal Drugs and New Animal Drug Combination Products Administered in or on Medicated Feed or Drinking Water of Food-Producing Animals: Recommendations for Drug Sponsors for Voluntarily Aligning Product Use Conditions with GFI #209 [PDF], which attempts to set up a framework for voluntary withdrawal and/or labeling changes for antimicrobials approved prior to 2003.
Unfortunately, the FDA's actions are likely to satisfy only a few stakeholders. Since it continues to adhere to voluntary adoption of limited production antibiotic use, it will make unhappy those people and industry participants who believe that formal regulation is necessary to move quickly enough and to force outliers into the framework (e.g., NGOs such as the plaintiffs in NRDC v. the FDA, or companies who refuse to use meat from antibiotic-treated animals). Similarly, since it recommends a system of voluntary withdrawals, it will not please those people who believe that mandatory withdrawals are necessary. In addition, the FDA's actions focus mostly on production use of antibiotics but there are people who believe that the problem of antibiotic resistance cannot be fully tackled unless treatment uses are also addressed. On the other hand, other industry participants are likely to assert that the FDA's actions will increase the costs of meat production by eliminating growth enhancement methods without sufficient evidence that production use actually causes antibiotic resistance. Some of the most unhappy industry participants are likely to be some small producers who are least equipped to deal with the added costs from increased veterinary oversight and the elimination of OTC drugs.
But even if the FDA's strategy is unlikely to satisfy many stakeholders, it may still be the correct strategy. It does attempt to address the main scientific findings about the contribution of animal antibiotics use to antibiotic resistance generally. It ties restrictions on use to the drugs that are most important for human health. That list is likely to evolve over time. Drugs approved since GFI #152 was issued in 2003 meet the judicious use standards. Requiring formal withdrawal proceedings for older approvals would severely tax administrative resources and would likely result in some litigation. Formal regulation may not provide enough flexibility for the FDA to respond to changing information and might itself result in litigation. If the FDA can achieve close-to-full compliance through voluntary means, that seems much more efficient. But the FDA needs to study whether that is indeed the case and better surveillance of compliance is needed. It cannot proceed on a mere hope and a promise.
Development of a full voluntary framework for the judicious use of antibiotics in animals is a central part of the FDA's Foods and Veterinary Medicine Strategic Plan [PDF] that was published in April. But the FDA is also correct to view antibiotic use in animals as only one piece in the puzzle of antibiotic resistance overall. Inappropriate human use and incentives for producing new and innovative antibiotics are at least equally important. Thus, this problem implicates many players: many other stakeholders, other centers within the FDA, other agencies within Health and Human Services, the USDA and many other national and international organizations. The risk, however, is that the breadth of the problem means that no actor will take full responsibility for its solution. We have wrestled with this problem for more than forty years. It is time that we have a comprehensive and effective plan to deal with it.
Margaret Riley is a Professor of Law at the University of Virginia School of Law. She serves as the chair of the University of Virginia's Embryonic Stem Cell Research Oversight Committee and as legal advisor to the Health Sciences Institutional Review Board. Her research focuses on food and drug law, health law, biotechnology law and bioethics.
Suggested citation: Margaret Riley, The Regulation of Antibiotic Use in Animal Agriculture, JURIST - Forum, May 26, 2012, http://jurist.org/forum/2012/05/margaret-riley-antibiotics-fda.php.
This article was prepared for publication by Caleb Pittman, head of JURIST's academic commentary service. Please direct any questions or comments to him at firstname.lastname@example.org